note: This module has been peer-reviewed, accepted, and sanctioned by the National Council of Professors of Educational Administration (NCPEA) as a scholarly contribution to the knowledge base in educational administration.
Meta-analytic Model
The meta-analytic model is based in a common research question. The meta-analytic model represents the analysis of a specic topic from multiple perspectives or vantage points for the purpose of comparing and contrasting ndings. For example, a question may be, What are the perceptions of superintendents and their school board presidents regarding the role of superintendents in 1000 Texas public school districts? If a team of three individuals are working on this particular question, then the rst dissertation may explore the question from a small district perspective; the second dissertation may explore the question from a mid-size district perspective, while the third dissertation may explore the question from the large urban district perspective. In this case, Chapters One (Introduction) and Two (Review of Literature) of the dissertation could be entirely dierent; however, the dissertations may all include a collaboratively-derived theoretical framework. Chapter Three (Methodology) must present the sampling plan in diering ways; however, all three dissertations might include the same information on instrumentation, research design, and procedures (all planned together). Chapter Four (Results) and Five (Discussion) would be individually authored and would present the ndings for the unique samples and the summary, conclusions, and recommendations.
It is recommended in this model that the chair or co-chairs be the same person(s) for the companion dissertations because that person(s) would have the knowledge of all the multiple dissertations to guide dierent perspectives.
May (1991), McConaghy (1991), and Nolt (1991) are examples of three companion dissertations that represent such a meta-analytic model. McConaghy's abstract follows and demonstrates the meta-analytic quality of the approach related to learning and teaching in the elementary school. His study reects the portion of the study that is related to teachers.
Statement of the problem. The study was designed to explore the following problem: Does the IALS approach, utilizing hands-on science based activities integrated with mathematics and language arts and emphasizing a cooperative learning-teaching strategy, enhance learning of science, mathematics, and writing skills among diverse groups of fourth grade students; and does this approach evoke interest by these students and their teachers?
Procedure. The study was conducted by a three-member research team. The team developed the IALS around the theme of magnetism, developed and organized the support materials, trained the teachers, col-lected and analyzed the student and teacher data, and reported the ndings in three companion dissertations.
Part 1 analyzed data related to student performance, Part 2 analyzed data pertaining to student responses, and Part 3 analyzed data related to teacher responses.
The population included 348 students from 16 fourth grade classes from three school districts. The design of the study included an experimental group receiving the IALS treatment (n = 248), a traditional group, receiving a textbook treatment (n = 67), and a comparison group receiving no specic treatment (n = 23).
The IALS consisted of six lessons taught in an 8 week period. A Student Activity Book containing the activities of the IALS was included to serve as a means of collecting student responses to those activities.
7This content is available online at <http://cnx.org/content/m14542/1.1/>.
Also included was a Teacher's Guide including related objectives, goals, materials, preparation, investigative procedures, and content. All materials were provided in a classroom kit.
Teachers were interviewed to determine their preparation for utilizing the IALS, involvement of students in the IALS approach, meeting of stated learning goals, and satisfaction with the cooperative learning-teaching strategy.
Major conclusions. Teachers were able to eectively utilize the IALS approach, but recommended ad-ditional time be provided for teaching the individual lessons. Teachers were able to involve students and meet the established learning goals. Cooperative learning-teaching was not clearly viewed as favorable or unfavorable. Additionally, the overall study determined that an IALS could be developed and eectively implemented by elementary teachers among diverse groups of fourth grade students. ftn*This study was con-ducted in a three-part eort by team members Patricia Nell May (pub.# 9134976), Robert M. McConaghy (pub.# 9134977), and Sally K. Nolt (pub.# 9134985). (¶1-6)
The abstract of one of the companions is also shared as follows so that the relationships of two of the three companions can be made by the reader. The other companion dissertation in this trio is related in similar fashion. May's (1991) study reects the student research component to the problem and her abstract follows.
Statement of the problem. The study was designed to explore the following problem: Does the IALS approach, utilizing hands-on science based activities integrated with mathematics and language arts and emphasizing a cooperative learning-teaching strategy, enhance learning of science, mathematics, and writing skills among diverse groups of fourth grade students; and does this approach evoke interest by these students and their teachers?
Procedure. The study was conducted by a three-member research team. The team developed the IALS around the theme of magnetism, developed and organized the support materials, trained the teachers, col-lected and analyzed the student and teacher data, and reported the ndings in three companion dissertations.
Part 1 analyzed data related to student performance, Part 2 analyzed data pertaining to student responses, and Part 3 analyzed data related to teacher responses.
The population included 348 students from 16 fourth grade classes from three school districts. The design of the study included an experimental group receiving the IALS treatment (n = 248), a traditional group, receiving a textbook treatment (n = 67), and a comparison group receiving no specic treatment (n
= 23). Student performance was examined through pretest to posttest mean dierences of students in the experimental, traditional, and comparison groups.
The IALS consisted of six lessons taught in a 8 week period. A Student Activity Book containing the activities of the IALS was included to serve as a means of collecting student responses to those activities.
Also included was a Teacher's Guide including related objectives, goals, materials, preparation, investigative procedures, and content. All materials were provided in a classroom kit.
Major conclusions. Students using the IALS approach experienced the greatest success in strengthening mathematics skills. Students using both the IALS and textbook approaches learned the intended science knowledge and writing skills.
The IALS approach resulted in signicant academic gains for all genders, ethnic groups, academic stand-ings, instructional programs, and learning preferences with the following exceptions: gains in science knowl-edge were comparatively lower for ESL students, as were gains in mathematics skills for the gifted. Ad-ditionally, the overall study determined that an IALS could be developed and eectively implemented by elementary teachers among diverse groups of fourth grade students. ftn*This study was conducted in a three-part eort by team members Patricia Nell May (pub.# 9134976), Robert M. McConaghy (pub.# 9134977), and Sally K. Nolt (pub.# 9134985). (¶1-6)
The two abstracts presented are the same with the exception of the conclusions section. Additionally, the abstracts of all three dissertations note in a footnote that the study is a part of a three-part companion dissertation. This should clearly be noted in the abstract, particularly as words are used that are the same, and to readers reading without this notation, it may appear as plagiarism. The note would stand to clarify that point. Recommendations for the misconception of plagiarism not to occur would be to alter each of the abstracts to be relevant to only that particular aspect of the meta-analytic study. Perhaps the method
and purpose may remain the same in general, but the specic aspect of the individual dissertation reported should be more prominent. Specify the larger sample, but hone in on the individual study's sample. The conclusions section is the only area in the abstract that make these three abstracts dierent to the public's review. More specicity to the individual studies is recommended in this type of meta-analytic companion dissertation as students collaborate and write their dissertations.
Multiple Case Study Model
A second format for the companion dissertation is the multiple case study model that may elaborate one or more research questions. The multiple case model for companion dissertations is characterized by a collaboratively developed research question that envelopes two, three, or more cases extending from that one question. For example, one research question for three companion dissertations could be, What content and instructional strategies should be included in a basic statistics course created for both practicing and prospective school principals? In this design, the three candidates would be expected to complete a one-year research seminar dedicated to the following essential tasks: (a) creating a common theoretical framework (review of relevant literature), (b) constructing a dissertation proposal that specied both common and individual research tasks for each candidate, and (c) designing a common framework for organizing the narrative to be presented in each companion dissertation. This format diers from the traditional ve chapter dissertation in which a separate review of literature chapter is presented as the traditional chapter two. In this model, the candidates' proposals would have four parts:
1. A collaborative elaboration of a theoretical framework developed in a research seminar during course-work.
2. A second section of the proposal including the common purpose and research questions to be addressed in each dissertation.
3. A third section of the proposal including the common research design used to guide the unique and collaborative empirical eorts to be undertaken in each companion dissertation.
4. A nal section of the proposal describing how the narrative for each dissertation would be organized into ve parts (chapters).
In this multiple case study model, all three dissertation proposals may be presented at on single proposal session. In this case, the same chair(s) and two committee members are recommended to be the same with a fourth committee member being dierent on each committee.
As an example, Jones (1999), Etheredge (1999), and Polnick (1999) developed three companion disserta-tions that followed an agreed-upon structure for a ve-chapter dissertation: (a) Introduction presented the theoretical framework and literature review, intent of inquiry (problem, purpose, and research questions;
basically the same chapter for all three dissertations), design of inquiry (process and expected outcome for each phase; basically the same chapter for all three), and a review of how the dissertation narrative was to be organized, (b) Phase One elaborated the intent, actual implementation method, and research ndings for the rst phase in which each candidate logged entries into a personal daily journal, took statements from school principal colleagues, and investigated appropriate situations encountered in the school principal literature dealing with data-based decision making (dissertation chapter diverges among the three), (c) Phase Two elaborated the intent, actual implementation method, and research ndings for the second phase in which the candidate served as a statistics and design consultant to a practicing principal during a data-based action research project (dissertation chapter diverges for each candidate), (d) Phase Three elaborated the intent, actual implementation method, and research ndings for the third phase in which each candidate developed an inventory of recommendations (dissertation is dierent for each), and (e) Summary and Conclusions pro-vided a brief overview of the detailed information presented in the rst four chapters (dissertation is similar yet divergent for each).
The abstracts of the multiple case study companion dissertations follow and are presented as the same abstract, though the dissertations dier by chapter as previously described. Polnick's abstract (same as Etheredge and Jones) follows.
The purpose of this inquiry was to test the feasibility of the McNamara and Thompson (1996) model for teaching statistics and data analysis methods in principal preparation programs.
Design. This inquiry was conducted in three phases. The rst phase was used to create an inventory of situations that describe opportunities for principals to use statistics and data analysis methods at the campus level. The second phase was a case study that required the author to assume a consultant role in conducting and reporting the results of a formative evaluation survey indicating the eectiveness of an ongoing sta development project aimed at improving test scores on a single middle school campus. The third phase was dedicated to constructing recommendations.
Findings. The rst phase identied 482 specic opportunities for data analysis. Most opportunities focused on instructional concerns and over 90 percent of these opportunities involved using descriptive rather than inferential statistics to analyze data already available on the campus. Completion of the formative evaluation case study in the second phase provided several additional insights about the data analysis skills principals need to be eective on the job.
Recommendations. Phase three provides twelve recommendations organized into three sections. The rst section provides two recommendations to conrm and endorse the guidelines advanced in the McNamara and Thompson (1996) model. Using the conrmatory evidence from the rst two phases, the second section oers seven specic recommendations for the design of a statistics course aimed at developing data analysis skills for prospective and practicing school principals. The third section presents three recommendations for continuing the research agenda initiated in this inquiry.
Evaluation Model
A third type of companion dissertation is considered the evaluation model. In such a model, the same question is asked for varying samples. An example of this model is exhibited by two candidates who desired to evaluate whether a highly focused Spanish early literacy intervention increased pre-reading skills for culturally and linguistically diverse students who were identied by their classroom teachers as most at risk for reading diculties. In these example companion dissertations, the evaluation agenda was the same, but the samples were dierent. One study was at the kindergarten level, and one was at the rst grade level.
Archival data were used. A common research design was used for both dissertations including the same statewide instrument and descriptive statistics. Chapters diered in both dissertations with the exception of the common evaluative agenda, purpose, and research questions. McArthur (2003) and Mohr (2003) authored their abstracts as follows.
McArthur (2003) reported her abstract in her dissertation in the following manner.
Purpose. The purpose of this exploratory study was to evaluate whether a highly focused Spanish early literacy intervention increased pre-reading skills for sixteen culturally and linguistically diverse bilingual kindergarten students who were identied as most at risk for reading diculties at the focus school. Two research questions were used to guide the empirical eorts of the study: (a) How eective was the intervention for improving student reading scores in each of the 12 instructional reading components addressed in the intervention? and (b) How eective was the intervention for each of the sixteen kindergarten bilingual students in the focus school's at risk population?
Design. A one-group pretest-posttest design was used to provide the empirical evidence needed to answer both research questions. The independent variable was the set of twelve unique reading intervention strategies that were developed for each of the twelve components measured by the Tejas LEE (TL), the common measure used in the pretest and posttest.
Results. For research question one, the Spanish early literacy intervention was determined to be eective for seven of the twelve TL literacy components addressed in the intervention.
For research question two, this Spanish early literacy intervention was determined to be "Highly Eec-tive" or "EecEec-tive" for eight kindergarten students, "Partially EecEec-tive" for an additional ve students and
"Ineective" for the remaining three kindergarten students.
Recommendations. Three recommendations were presented for continuing the research agenda addressed in this study. The initial recommendation suggested a need for researchers to reexamine the validity of the First Letter Omission measure used in the kindergarten TL. The other two recommendations indicate how and why practitioners should replicate this study for other at risk bilingual kindergarten students. (p.iii-iv) Mohr's dissertation abstract is clearly similar to McArthur's as it should be in a companion dissertation;
however, dierences are also clear in the abstract. Mohr's abstract follows.
Purpose. The intent of this inquiry was to determine the extent to which a highly focused Spanish early literacy intervention increased pre-reading skills for 13 culturally and linguistically diverse bilingual rst grade students who were identied in their school as most at risk for reading diculties. Two research questions were used to structure the empirical eorts of this inquiry:(a) How eective was the intervention for improving student reading scores in each of the 17 instructional reading components addressed in the intervention? and (b) How eective was the intervention for each of the 13 rst grade bilingual students designated as the at risk population?
Design. A single-group pretest-posttest design was used to generate the empirical evidence needed to answer these two research questions. The independent variable was the set of 17 unique reading intervention strategies that were developed for each of the 17 components measured in the rst grade version of the Tejas LEE (TL), the common measure used in the pretest and the posttest.
Question one ndings. Using specied decision rules for interpreting the pretest and posttest TL results, the Spanish early literacy intervention was determined to be eective for 15 of the 17 TL literacy components addressed in the intervention.
Question two ndings. This Spanish early literacy intervention was determined to be "Highly Eective" or
"Eective" for ten rst grade students and "Partially Eective" for the remaining three rst grade students.
Recommendations. Three recommendations were advanced for continuing the research agenda initiated in this inquiry. The rst two recommendations focused on future research aimed at improving the TL. The rst research recommendation called for developing a parallel form of the TL to eliminate the problem of test wiseness. The second research recommendation suggested a need to reexamine the validity of the Fluency and Reading Accuracy TL measures. The nal recommendation proposes what practitioners should do to replicate this study for the same intervention population. (p. iii-iv)
As indicated in the abstract of McArthur, the purpose was to evaluate whether a highly focused Spanish early literacy intervention increased pre-reading skills for sixteen culturally and linguistically diverse bilingual kindergarten students who were identied as most at risk for reading diculties at the focus school (p.
iii). Mohr addressed the similar purpose in this way for these companion dissertations, to determine the extent to which a highly focused Spanish early literacy intervention increased pre-reading skills for 13 culturally and linguistically diverse bilingual rst grade students who were identied in their school as most at risk for reading diculties (p.iii). The evaluation model for companion dissertations focuses on the same topic/issue/concern being evaluated as evidence in the referenced cases of McArthur and Mohr on the evaluation of an intervention in early Spanish literacy.
Single Case Model
A fourth type of companion dissertation is a single case model in which one context or setting is selected for the study; however, diering target populations are studied within that context. For example, one companion dissertation (Creel, 2000) sought to determine dierences in student achievement and attendance of African American students before and after the implementation of a standardized dress code in a suburban high school in Southeast Texas, while the other companion studied the same among Hispanic high school students (Widener, 2000). As part of the companion studies, dierences in student attendance between African American students and Hispanic students were studied to determine if dierences existed in two separate subpopulations in the same high school.
A fourth type of companion dissertation is a single case model in which one context or setting is selected for the study; however, diering target populations are studied within that context. For example, one companion dissertation (Creel, 2000) sought to determine dierences in student achievement and attendance of African American students before and after the implementation of a standardized dress code in a suburban high school in Southeast Texas, while the other companion studied the same among Hispanic high school students (Widener, 2000). As part of the companion studies, dierences in student attendance between African American students and Hispanic students were studied to determine if dierences existed in two separate subpopulations in the same high school.